Yamaha has just put a humanoid robot on one of its motorcycles and pitted it against Valentino Rossi, a professional rider. The robot didn’t win, but it’s amazing that it could ride at all.
Feast your eyes:
What’s striking is that the bike is unmodified: the robot is a hunched-over form on top. It senses the environment, calculates what to do, keeps the bike stable, manages acceleration and deceleration—all while factoring in road conditions, air resistance, and engine braking.
The project is “a moon shot,” says Hiroshi Saijo, the chief executive of Yamaha. They chose to do it because it is hard. But it’s also part of an industry-wide drive to incorporate autonomous features into commercial motorcycles to make them easier and safer to ride.
For instance, Yamaha’s bike can remain stable at speeds as low as 15 kph (9 mph). But that’s nothing compared to Honda’s latest feat, an all-electric concept called Riding Assist-e, which can stand up when moving at a snail’s pace and even when at a dead stop. It works rather like a Segway, balancing itself with tiny motions of the wheels. Kawasaki is taking another route by developing an AI voice-response system similar to Siri, both to help the driver stay abreast of the situation and to let the AI monitor the driver’s attentiveness.
Yamaha’s Motobot, by contrast, didn’t have be safe or user-friendly. There’s no human rider, and the bike’s only purpose is to go fast. In a straight line, it can hit 200 kph (124 mph). When rounding the track, though, its lap time came to 117.50 seconds—nearly 32 seconds short of Rossi’s best, 85.74 seconds.
Wait ‘til next time.
Unlike self-driving car prototypes, Motobot doesn’t ever navigate unknown roads and it never encounters traffic, pedestrians, or stray dogs. It merely needs to have a sense of where it is on a pre-mapped track.
It can locate itself to within 2.5 centimeters using a combination of inertial measurement and GPS-RTK—GPS with real-time kinematics. That means augmenting GPS by referring to a signal from a radio station. Motobot also monitors the bike’s acceleration, air resistance, engine braking, and road conditions.
Yamaha’s engineers simulated all these elements in parallel, first on a computer and then in a hybrid test bed that couples a computer to the vehicle as it rolls on a dynamometer. That’s the machine used to physically test a car while it's rolling against a load but sitting stationary in a lab.
Philip E. Ross is a senior editor at IEEE Spectrum. His interests include transportation, energy storage, AI, and the economic aspects of technology. He has a master's degree in international affairs from Columbia University and another, in journalism, from the University of Michigan.